Search NASA⌕ Search

SEARCH · Search NASA

Results for “gradient methods”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,279 records · Page 71

ZnTeO{sub 3} crystal growth by a modified Bridgman technique

Highlights: • ZnTeO{sub 3} single crystals were grown for the first time by a modified Bridgman method. • The growth is still possible in a system that lacks congruent melting. • A growth is best when melt is exposed to a steeper axial thermal gradient. • Optical and electrical properties were investigated for the grown crystals. - Abstract: Zinc Tellurite (ZnTeO{sub 3}) crystals were grown for the first time using a modified Bridgman method with a 2.5 kHz radio frequency (RF) furnace. Single crystal growth of ZnTeO{sub 3} was hindered by many complicating factors, such as the evaporation of TeO{sub 2} above 700 °C and the formation of more than one phase during crystal growth. While there were several successful runs that produced ZnTeO{sub 3} single crystals, it was found that large (≥10 cm{sup 3}) single ZnTeO{sub 3} crystals resulted when the crucible was exposed to a steeper vertical thermal gradient and when the temperature of the melt was raised to at least 860 °C. The results of powder X-ray diffraction (XRD) patterns were in accordance with the X-ray powder diffraction file (PDF) for ZnTeO{sub 3}. Some optical, electrical and structural properties of ZnTeO{sub 3} single crystals were reported in this paper.

36 MATERIALS SCIENCE↗

A time-parallel multiple-shooting method for large-scale quantum optimal control

Quantum optimal control plays a crucial role in quantum computing by providing the interface between compiler and hardware. Solving the optimal control problem is particularly challenging for multi-qubit gates, due to the exponential growth in computational complexity with the system's dimensionality and the deterioration of optimization convergence. To ameliorate the computational complexity of time-integration, this paper introduces a multiple-shooting approach in which the time domain is divided into multiple windows and the intermediate states at window boundaries are treated as additional optimization variables. Further, this enables parallel computation of state evolution across time-windows, significantly accelerating objective function and gradient evaluations. Since the initial state matrix in each window is only guaranteed to be unitary upon convergence of the optimization algorithm, the conventional gate trace infidelity is replaced by a generalized infidelity that is convex for non-unitary state matrices. Continuity of the state across window boundaries is enforced by equality constraints. A quadratic penalty optimization method is used to solve the constrained optimal control problem, and an efficient adjoint technique is employed to calculate the gradients in each iteration. We demonstrate the effectiveness of the proposed method through numerical experiments on quantum Fourier transform gates in systems with 2, 3, and 4 qubits, noting a speedup of 80x for evaluating the gradient in the 4-qubit case, highlighting the method's potential for optimizing control pulses in multi-qubit quantum systems.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Calibration of EBT2 film using a red-channel PDD method in combination with a modified three-channel technique

Purpose: Ashland Inc. EBT2 and EBT3 films are widely used in quality assurance for radiation therapy; however, there remains a relatively high degree of uncertainty [B. Hartmann, M. Martisikova, and O. Jakel, “Homogeneity of Gafchromic EBT2 film,” Med. Phys. 37, 1753–1756 (2010)]. Micke et al. (2011) recently improved the spatial homogeneity using all color channels of a flatbed scanner; however, van Hoof et al. (2012) pointed out that the corrected nonuniformity still requires further investigation for larger fields. To reduce the calibration errors and the uncertainty, the authors propose a new red-channel percentage-depth-dose method in combination with a modified three-channel technique. Methods: For the ease of comparison, the EBT2 film image used in the authors’ previous study (2012) was reanalyzed using different approaches. Photon beams of 6-MV were delivered to two different films at two different beam on times, resulting in the absorption doses of ranging from approximately 30 to 300 cGy at the vertical midline of the film, which was set to be coincident with the central axis of the beam. The film was tightly sandwiched in a 30{sup 3}-cm{sup 3} polystyrene phantom, and the pixel values for red, green, and blue channels were extracted from 234 points on the central axis of the beam and compared with the corresponding depth doses. The film was first calibrated using the multichannel method proposed by Micke et al. (2010), accounting for nonuniformities in the scanner. After eliminating the scanner and dose-independent nonuniformities, the film was recalibrated via the dose-dependent optical density of the red channel and fitted to a power function. This calibration was verified via comparisons of the dose profiles extracted from the films, where three were exposed to a 60° physical wedge field and three were exposed to composite fields, and all of which were measured in a water phantom. A correction for optical attenuation was implemented, and treatment plans of intensity modulated radiation therapy and volumetric modulated arc therapy were evaluated. Results: The method described here demonstrated improved accuracy with reduced uncertainty. The relative error compared with the measurements of a water phantom was less than 1%, and the overall calibration uncertainty was less than 2%. Verification tests revealed that the results were close to those of the authors’ previous study, and all differences were within 3%, except those with a high-dose gradient. The gamma pass rates (2%/2 mm) of the treatment plan evaluated using the method described here were greater than 99%, and no obvious stripe patterns were observed in the dose-difference maps. Conclusions: Spatial homogeneity was significantly improved via the calibration method described here. This technique is both convenient and time-efficient because it does not require cutting the film, and only two exposures are necessary.

60 APPLIED LIFE SCIENCES↗

Monte Carlo Simulations of Coupled Diffusion and Surface Reactions during the Aqueous Corrosion of Borosilicate Glasses

Borosilicate nuclear waste glasses develop complex altered layers as a result of coupled processes such as hydrolysis of network species, condensation of Si species, and diffusion. However, diffusion has often been overlooked in Monte Carlo models of the aqueous corrosion of borosilicate glasses. Therefore, three different models for dissolved Si diffusion in the altered layer were implemented in a Monte Carlo model and evaluated for glasses in the compositional range (75-x) mol% SiO2 (12.5+x/2) mol% B2O3 and (12.5+x/2) mol% Na2O, where 0 ≤ x ≤ 20%, and corroded in static conditions at a surface-to-volume ratio of 1000 m-1. The three models considered instantaneous homogenization (M1), linear concentration gradients (M2), and concentration profiles determined by solving Fick’s 2nd law using a finite difference method (M3). Model M3 revealed that concentration profiles in the altered layer are not linear and show changes in shape and magnitude as corrosion progresses, unlike those assumed in model M2. Furthermore, model M3 showed that, for borosilicate glasses with a high forward dissolution rate compared to the diffusion rate, the gradual polymerization and densification of the altered layer is significantly delayed compared to models M1 and M2. Models M1 and M2 were found to be appropriate models only for glasses with high release rates such as simple borosilicate glasses with low ZrO2 content.

borosilicates↗

Conceptual design for a blanket tritium extraction test stand

Tritium breeding is fundamentally required for a sustainable fusion fuel cycle, yet the technological readiness of blanket technology lags far behind other fusion systems. Breeder concepts are divided into solid and liquid media, where solid breeders typically rely on a sweep gas, such as helium, to carry away tritium from lithium containing ceramic materials, and liquid breeders produce tritium from lithium containing eutectics (e.g., PbLi) or molten salts (FLiBe). In each case, tritium must be harvested from the breeding medium. A promising method for tritium extraction is through a vacuum permeator, in which a concentration gradient from the tritium- containing fluid promotes diffusion through a membrane with high hydrogen permeability to the vacuum. This technology has been demonstrated for hydrogen gas systems using Pd and PdAg permeators, but relatively little work has been done to test tritium extraction from PbLi. A Tritium Extraction eXperimental (TEX) loop is being designed to test tritium extraction in a vacuum permeator configuration. The system design is such that it will allow the testing of tritium extraction from both helium and PbLi. A phased approach is being taken that will allow testing of small specimens for fundamental permeation measurements, to multi-meter component testing at near-prototypic conditions. A molten PbLi loop is challenging due to the toxic and explosive nature of Pb and Li, respectively, radiological concerns by introducing tritium, and high temperatures involved in such a system. In addition, PbLi corrosion is a significant issue at high temperatures (>400C). The TEX system will not employ a neutron source for volumetric production of tritium. Herein we present the design and methods used for 1) pumping PbLi, 2) introducing deuterium or tritium into the PbLi, 3) quantifying the amount of deuterium in the loop, 4) extracting deuterium and tritium, and 5) quantifying the total amount of extracted deuterium or tritium from the permeator. In addition, the safety design for operating such a system will be discussed.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Analytic energy gradient of excited electronic state within TDDFT/MMpol framework: Benchmark tests and parallel implementation

The time-dependent density functional theory (TDDFT) has become the most popular method to calculate the electronic excitation energies, describe the excited-state properties, and perform the excited-state geometric optimization of medium and large-size molecules due to the implementation of analytic excited-state energy gradient and Hessian in many electronic structure software packages. To describe the molecules in condensed phase, one usually adopts the computationally efficient hybrid Quantum Mechanics/Molecular Mechanics (QM/MM) models. Here, we extend our previous work on the energy gradient of TDDFT/MM excited state to account for the mutual polarization effects between QM and MM regions, which is believed to hold a crucial position in the potential energy surface of molecular systems when the photoexcitation-induced charge rearrangement in the QM region is drastic. The implementation of a simple polarizable TDDFT/MM (TDDFT/MMpol) model in Q-Chem/CHARMM interface with both the linear response and the state-specific features has been realized. Several benchmark tests and preliminary applications are exhibited to confirm our implementation and assess the effects of different treatment of environmental polarization on the excited-state properties, and the efficiency of parallel implementation is demonstrated as well.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Achieving band gap grading of CZTS and CZTSe materials

Techniques for achieving band gap grading in CZTS/Se absorber materials are provided. In one aspect, a method for creating band gap grading in a CZTS/Se absorber layer includes the steps of: providing a reservoir material containing Si or Ge; forming the CZTS/Se absorber layer on the reservoir material; and annealing the reservoir material and the CZTS/Se absorber layer under conditions sufficient to diffuse Si or Ge atoms from the reservoir material into the CZTS/Se absorber layer with a concentration gradient to create band gap grading in the CZTS/Se absorber layer. A photovoltaic device and method of forming the photovoltaic device are also provided.

36 MATERIALS SCIENCE↗

Optimization of a Method To Quantify Soil Bacterial Abundance by Flow Cytometry

ABSTRACT Bacterial abundance is a fundamental metric for understanding the population dynamics of soil bacteria and their role in biogeochemical cycles. Despite its importance, methodological constraints hamper our ability to assess bacterial abundance in terrestrial environments. Here, we aimed to optimize the use of flow cytometry (FCM) to assay bacterial abundances in soil while providing a rigorous quantification of its limitations. Soil samples were spiked with Escherichia coli to evaluate the levels of recovery efficiency among three extraction approaches. The optimized method added a surfactant (a tetrasodium pyrophosphate [TSP] buffer) to 0.1 g of soil, applied an intermediate degree of agitation through shaking, and used a Nycodenz density gradient to separate the cells from background debris. This procedure resulted in a high (average, 89%) level of cell recovery. Recovery efficiencies did not differ significantly among sites across an elevation gradient but were positively correlated with percent carbon in the soil samples. Estimated abundances were also highly repeatable between technical replicates. The method was applied to samples from two field studies and, in both cases, was sensitive enough to detect treatment and site differences in bacterial abundances. We conclude that FCM offers a fast and sensitive method to assay soil bacterial abundance from relatively small amounts of soil. Further work is needed to assay differential biases of the method across a wider range of soil types. IMPORTANCE The ability to quantify bacterial abundance is important for understanding the contributions of microbial communities in soils, but such assays remain difficult and time-consuming. Flow cytometry offers a fast and direct way to count bacterial cells, but several concerns remain in applying the technique to soils. This study aimed to improve the efficiency of the method for soil while quantifying its limitations. We demonstrated that an optimized procedure was sensitive enough to capture differences in bacterial abundances among treatments and ecosystems in two field studies.

54 ENVIRONMENTAL SCIENCES↗

High-throughput fabrication and semi-automated characterization of oxide thin film transistors

High throughput experimental methods are known to accelerate the rate of research, development, and deployment of electronic materials. For example, thin films with lateral gradients in composition, thickness, or other parameters have been used alongside spatiallyresolved characterization to assess how various physical factors affect material properties under varying measurement conditions. Similarly, multi-layer electronic devices that contain such graded thin films as one or more of their layers can also be characterized spatially in order to optimize the performance. In this work, we apply these high throughput experimental methods to thin film transistors (TFTs), demonstrating combinatorial device fabrication and semiautomated characterization using sputtered oxide TFTs as a case study. Here, we show that both extrinsic and intrinsic types of device gradients can be generated in a TFT library, such as channel thickness and length, channel cation compositions, and oxygen atmosphere during deposition. We also present a semi-automated method to measure the 44 devices fabricated on a 50x50mm substrate that can help to identify properly functioning TFTs in the library and finish the measurement in a short time. Finally, we propose a fully automated characterization system for similar TFT libraries, which can be coupled with high throughput data analysis. These results demonstrate that high throughput methods can accelerate the investigation of TFTs and other electronic devices.

36 MATERIALS SCIENCE↗

Gradient-Descent-Based Learning Gain for Backstepping Controller and Disturbance Observer of Nonlinear Systems

This paper proposes a gradient-descent-based learning (GL) gain for backstepping controller and disturbance observer (DOB) of nonlinear system. The proposed method consists of the GL gain update law, controller, and DOB. The GL gain update law is proposed to adapt the control gain and DOB gain according to the direction that minimizes the cost function. The mathematical analysis reveals that the GL gain always has a positive sign and upper bound. The controller is designed via a backstepping procedure to track the desired output with GL control gain. The DOB is designed to estimate the unknown external disturbance with the GL DOB gain. Because the control and DOB gains are simultaneously tuned to achieve improved performance, the time consumption for tuning can be reduced. In addition, the peaking phenomenon can be avoided initially by a small initial value of GL gains. The stability of the closed-loop system is guaranteed using the input-to-state stability property. The performance of the proposed method was validated via simulations and experiments using a DC motor.

42 ENGINEERING↗

Correlating Surface Processing of Nb Superconducting RF Cavities with the Evolution of Surface Electronic States

Superconducting-radio frequency (RF) cavities provide an efficient way to accelerate particle beams with extremely high acceleration gradients while generating very small power dissipation. The few nanometers of the surface play a critical role in defining the RF performance of superconducting Nb based cavities. Over the past two decades, several pioneering surface treatment and processing methods have emerged, enabling remarkable improvements in superconducting cavity performance by simultaneously achieving high quality factors with increasing maximum acceleration gradients. These processing approaches include chemical polishing, distinct multi-step thermal treatments under ultra-high vacuum (UHV) conditions over low to high temperature regimes, as well as high-temperature treatments under controlled nitrogen atmospheres. Beyond their macroscopic impact on RF performance, these methods produce distinct surface oxide configurations characterized by different valence states, oxide thicknesses, chemical uniformity, and oxygen concentration profiles extending into the near-surface bulk of niobium. In this work, we are trying to understand how the surface-processing methods and the resulting oxide/oxygen profiles affect the electronic structure of surface and the mechanism of superconductivity. Using a combination of X-ray photoemission and X-ray absorption spectroscopies, we investigate how the valence-band structure and the electronic density of states (DoS) near the Fermi level evolve under different surface treatments. By employing tunable photon energies across multiple elemental absorption edges, we use resonant photoemission to disentangle and identify the elemental contributions to specific valence-band features. Our observations show that different surface-processing methods lead to distinct temperature evolutions of the DoS and valence-band states near the Fermi level. Our results suggest variations in different Nb-O orbital hybridizations in distinct processes and point towards the possibility of different underlying mechanisms of superconductivity governed by surface chemistry and oxide configuration. We also correlate these distinct superconducting mechanisms with RF cavity performance, specifically focusing on measured surface resistance, the nature of the Q-slope, and quench fields observed in superconducting RF measurements. These results will enable us to identify the potential limiting factors and relevant controllable parameters that can be further optimized to improve the performance of superconducting RF cavities.

Tripathi, Malvika [Fermilab] (ORCID:00000001989251↗

Layer-Parallel Training of Deep Residual Neural Networks

Residual neural networks (ResNets) are a promising class of deep neural networks that have shown excellent performance for a number of learning tasks, e.g., image classification and recognition. Mathematically, ResNet architectures can be interpreted as forward Euler discretizations of a nonlinear initial value problem whose time-dependent control variables represent the weights of the neural network. Hence, training a ResNet can be cast as an optimal control problem of the associated dynamical system. For similar time-dependent optimal control problems arising in engineering applications, parallel-in-time methods have shown notable improvements in scalability. This paper demonstrates the use of those techniques for efficient and effective training of ResNets. The proposed algorithms replace the classical (sequential) forward and backward propagation through the network layers with a parallel nonlinear multigrid iteration applied to the layer domain. This adds a new dimension of parallelism across layers that is attractive when training very deep networks. From this basic idea, we derive multiple layer-parallel methods. The most efficient version employs a simultaneous optimization approach where updates to the network parameters are based on inexact gradient information in order to speed up the training process. Finally, using numerical examples from supervised classification, we demonstrate that the new approach achieves a training performance similar to that of traditional methods, but enables layer-parallelism and thus provides speedup over layer-serial methods through greater concurrency.

97 MATHEMATICS AND COMPUTING↗

Atomic-scale thermocapillary flow in focused ion beam milling

Focused ion beams provide a means of nanometer-scale manufacturing and material processing, which is used for applications such as forming nanometer-scale pores in thin films for DNA sequencing. We investigate such a configuration with Ga{sup +} bombardment of a Si thin-film target using molecular dynamics simulation. For a range of ion intensities in a realistic configuration, a recirculating melt region develops, which is seen to flow with a symmetrical pattern, counter to how it would flow were it driven by the ion momentum flux. Such flow is potentially important for the shape and composition of the formed structures. Relevant stress scales and estimated physical properties of silicon under these extreme conditions support the importance thermocapillary effects. A flow model with Marangoni forcing, based upon the temperature gradient and geometry from the atomistic simulation, indeed reproduces the flow and thus could be used to anticipate such flows and their influence in applications.

36 MATERIALS SCIENCE↗

Stochastic gradient descent algorithm for stochastic optimization in solving analytic continuation problems

We propose a stochastic gradient descent based optimization algorithm to solve the analytic continuation problem in which we extract real frequency spectra from imaginary time Quantum Monte Carlo data. The procedure of analytic continuation is an ill-posed inverse problem which is usually solved by regularized optimization methods, such like the Maximum Entropy method, or stochastic optimization methods. The main contribution of this work is to improve the performance of stochastic optimization approaches by introducing a supervised stochastic gradient descent algorithm to solve a flipped inverse system which processes the random solutions obtained by a type of Fast and Efficient Stochastic Optimization Method.

97 MATHEMATICS AND COMPUTING↗

Electric Gradient Breakdown Results for the Orthogonal Box Cavity in a 3 T Magnetic Field

The Muon Collaboration (Palmer and D. Stratakis) has developed a theory for A method to suppress high electric field breakdown in vacuum cavities in high magnetic fields that are needed for a muon collider or neutrino factory. It has been shown in our studies and by others that high gradient electric field emitted electrons (dark current) are the primary cause of breakdown. A DC magnetic field orthogonal to the RF electric accelerating field prevents dark current high field emitted electrons from traveling across the accelerating gap and therefore will prevent breakdown. We have built a test cavity ( TM‐2473‐APC) in the shape of a box to test this theory . The results of the first study are presented.

43 PARTICLE ACCELERATORS↗

Sheath expansion and plasma dynamics in the presence of electrode evaporation: Application to a vacuum circuit breaker

During the postarc dielectric recovery phase in a vacuum circuit breaker, a cathode sheath forms and expels the plasma from the electrode gap. The success or failure of current breaking depends on how efficiently the plasma is expelled from the electrode gap. The sheath expansion in the postarc phase can be compared to sheath expansion in plasma immersion ion implantation except that collisions between charged particles and atoms generated by electrode evaporation may become important in a vacuum circuit breaker. In this paper, we show that electrode evaporation plays a significant role in the dynamics of the sheath expansion in this context not only because charged particle transport is no longer collisionless but also because the neutral flow due to evaporation and temperature gradients may push the plasma toward one of the electrodes. Using a hybrid model of the nonequilibrium postarc plasma and cathode sheath coupled with a direct simulation Monte Carlo method to describe collisions between heavy species, we present a parametric study of the sheath and plasma dynamics and of the time needed for the sheath to expel the plasma from the gap for different values of plasma density and electrode temperatures at the beginning of the postarc phase. This work constitutes a preliminary step toward understanding and quantifying the risk of current breaking failure of a vacuum arc.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Selective-diffusion regularization for enhancement of microcalcifications in digital breast tomosynthesis reconstruction

Purpose: Digital breast tomosynthesis (DBT) has been shown to improve mass detection. Detection of microcalcifications is more challenging because of the large breast volume to be searched for subtle signals. The simultaneous algebraic reconstruction technique (SART) was found to provide good image quality for DBT, but the image noise is amplified with an increasing number of iterations. In this study, the authors developed a selective-diffusion (SD) method for noise regularization with SART to improve the contrast-to-noise ratio (CNR) of microcalcifications in the DBT slices for human or machine detection. Methods: The SD method regularizes SART reconstruction during updating with each projection view. Potential microcalcifications are differentiated from the noisy background by estimating the local gradient information. Different degrees of regularization are applied to the signal or noise classes, such that the microcalcifications will be enhanced while the noise is suppressed. The new SD method was compared to several current methods, including the quadratic Laplacian (QL) method, the total variation (TV) method, and the nonconvex total p-variation (TpV) method for noise regularization with SART. A GE GEN2 prototype DBT system with a stationary digital detector was used for the acquisition of DBT scans at 21 angles in 3 deg. increments over a {+-}30 deg. range. The reconstruction image quality without regularization and that with the different regularization methods were compared using the DBT scans of an American College of Radiology phantom and a human subject. The CNR and the full width at half maximum (FWHM) of the line profiles of microcalcifications within the in-focus DBT slices were used as image quality measures. Results: For the comparison of large microcalcifications in the DBT data of the subject, the SD method resulted in comparable CNR to the nonconvex TpV method. Both of them performed better than the other two methods. For subtle microcalcifications, the SD method was superior to other methods in terms of CNR. In both the subject and phantom DBT data, for large microcalcifications, the FWHM of the SD method was comparable to that without regularization, which was wider than that of the TV type methods. For subtle microcalcifications, the SD method had comparable FWHM values to the TV type methods. All three regularization methods were superior to the QL method in terms of FWHM. Conclusions: The SART regularized by the selective-diffusion method enhanced the CNR and preserved the sharpness of microcalcifications. In comparison with three existing regularization methods, the selective-diffusion regularization was superior to the other methods for subtle microcalcifications.

60 APPLIED LIFE SCIENCES↗

Modelling of the AGS using Zgoubi - Status

This paper summarizes the progress achieved so far, and discusses various outcomes, regarding the development of a model of the Alternating Gradient Synchrotron at the RHIC collider. The model, based on stepwise ray-tracing methods, includes beam and polarization dynamics. This is an on-going work, and a follow-on of code developments and particle and spin dynamics simulations that have been subject to earlier publications at IPAC and PAC [1, 2, 3]. A companion paper [4] gives additional informations, regarding the use of the measured magnetic field maps of the AGS main magnets.

43 PARTICLE ACCELERATORS↗